Breaking Down Barriers in Science

The world of science has undergone significant changes over the past 15 years, and Kaiming He, the Douglas Ross (1954) Career Development Professor of Software Technology in the Department of Electrical Engineering and Computer Science at MIT, has witnessed it firsthand. As a PhD student, He recalls a time when different disciplines and subjects were separated by high walls, making it challenging for researchers to understand and collaborate with each other.

A Common Language Emerges

However, with the advent of artificial intelligence (AI), particularly the "deep learning revolution" in 2012, these walls began to crumble. AI has created a common language that enables researchers from diverse fields to communicate and work together more effectively. He notes that this shift has been "very rare in human scientific history" and has the potential to break down barriers between different scientific disciplines.

The Power of AI

The deep learning revolution started with computer vision, which involves helping computers to "see" and perceive the world like humans. This technology has grown rapidly, and its applications have expanded to other areas of computer science, such as natural language processing, speech recognition, and robotics. The development of AI tools like ChatGPT and the progress toward artificial general intelligence (AGI) have been significant outcomes of this trend.

AI in Other Scientific Disciplines

One notable example of AI’s impact on other scientific disciplines is AlphaFold, a program developed by Google DeepMind that predicts protein structure. This application of AI in a different scientific discipline demonstrates the potential for AI to propagate and solve problems in various fields. He believes that this is just the beginning and that AI will continue to emerge in other areas of science.

Collaborations and Breakthroughs

Since joining MIT, He has had the opportunity to interact with professors from almost every department. Despite their diverse backgrounds, they can discuss deep learning, machine learning, and neural network models, which serve as a common language. This exchange of ideas has the potential to lead to significant breakthroughs and collaborations. For instance, AI can be used to analyze videos to predict weather and climate trends or to expedite the research cycle and reduce costs in new drug discovery.

The Future of AI in Science

The integration of AI into various scientific disciplines has significant potential for growth and innovation. While AI tools can benefit scientists, the reciprocal effect is also crucial, as scientists provide new problems and challenges that help evolve AI tools. He notes that many AI tools stem from earlier scientific areas, such as artificial neural networks, which were inspired by biological observations.

A New Era of Collaboration

The MIT Schwarzman College of Computing has created an environment that fosters connections between people from different disciplines, allowing them to share ideas and work together. He believes that this is an ideal place for scientists and AI researchers to come together and speak a common language. Although the complete lowering of barriers will take time, He expects that in 10 or more years, everyone will be using AI in some way for their research, just as computers are now a basic tool for everyone.

Conclusion

In conclusion, the emergence of AI has created a common language that is breaking down barriers between different scientific disciplines. As researchers like Kaiming He continue to collaborate and share ideas, the potential for innovation and growth is vast. The future of AI in science is exciting, and it will be interesting to see how these developments unfold in the years to come. With the continued advancement of AI, we can expect to see significant breakthroughs and collaborations that will shape the future of science and beyond.